CAREER: Towards the Next Generation of Data-Driven and Performance-Based Multiscale Procedures in Mining Geotechnics

该学院早期职业发展(CALEAR)奖将为在采矿岩土技术中创建数据驱动、物理指导和基于性能的多尺度程序奠定基础。该项目利用新兴数据科学领域提供的前所未有的机遇，重塑采矿岩土技术领域，并转变高风险基础设施的评估范式，重点放在尾矿储存设施上。在过去的十年中，尾矿储存设施的故障在世界范围内造成了前所未有的环境后果和生命损失。美国的一次失败，与最近的其他失败类似，可能会对环境、州经济和当地社区造成巨大的破坏。因此，尾矿存储设施的弹性设计和状况评估对于采矿活跃的地区至关重要，如亚利桑那州、内华达州、科罗拉多州、犹他州等，考虑到新的全球尾矿标准，尾矿存储设施的弹性设计和状况评估比以往任何时候都更加重要。在这方面，该项目将为提高尾矿储存设施的复原力提供基本见解，并为提高采矿基础设施的复原力设定新的标准。综合教育计划将通过在(I)佐治亚理工学院STEM中心和促进采矿基础设施恢复能力的美国机构；(Ii)佐治亚理工学院垂直综合项目；以及(Iii)教育增强机会，吸引本科生、研究生和K-12学生，帮助培养具有数据科学素养的下一代岩土尾矿工程师。外展计划将为拉美裔/拉丁裔学生建立一个同行试点辅导计划，以补充PI目前的外展活动。该项目的研究目标是(I)调查尾矿的多尺度机械响应，并通过数据科学发现高维相互作用；(Ii)调查现场状态评估的基本问题；以及(Iii)探索在采矿岩土工程中制定新的数据驱动和基于性能的程序。为了实现这些目标，本项目采用综合方法，综合考虑了数据科学、实验室测试、微结构测量、现场测试和数值模拟，对尾矿等中间材料的多尺度响应提出了新的见解，考虑了(I)微结构特征；(Ii)固有和诱发的各向异性和循环载荷的作用；(Iii)现场尺度的系统效应；以及(Iv)数据驱动的状态反演方法。从这项研究中获得的见解将被用来探索数据科学的价值，并正式使用它来解开尾矿力学特性的高维相互作用，并使用机器和主动学习来制定新的数据驱动和基于性能的程序。该项目将允许PI为数据科学、采矿岩土技术和基于性能的工程相结合的跨学科领域奠定基础，从而向岩土和灾害社区展示采用数据驱动方法的机会。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This Faculty Early Career Development (CAREER) award will set the stage for creating data-driven, physics-guided, and performance-based multiscale procedures in mining geotechnics. This project capitalizes on the unprecedented opportunities provided by the emerging field of data science to reshape the field of mining geotechnics and shift the paradigms for the assessment of high-risk infrastructure, focusing on tailings storage facilities. During the last decade, tailings storage facility failures have caused unprecedented environmental consequences and loss of human lives worldwide. A failure in the United States, similar to other recent ones, may cause dramatic damage to the environment, state economies, and local communities. Therefore, the resilient design and condition assessment of tailings storage facilities are vitally important for regions where mining is active, such as Arizona, Nevada, Colorado, Utah, amongst others, and have become more relevant than ever, considering the new global tailings standards. In this context, this project will provide fundamental insights to improve the resilience of tailings storage facilities and set new standards to enhance the resilience of mining infrastructure in general. The integrated educational plan will contribute to creating the next generation of geotechnical tailings engineers with literacy in data science by engaging undergraduate, graduate, and K-12 students in (i) Georgia Tech STEM centers and US institutions that promote the resilience of mining infrastructure; (ii) Georgia Tech vertically integrated projects; and (iii) education enhancement opportunities. The outreach plan will establish a peer pilot mentoring program for Hispanic/Latinx students that complements the PI’s current outreach activities.The research objectives of this project are to (i) investigate the multiscale mechanical response of mine tailings and discover high-dimensional interactions through data science; (ii) investigate the fundamental problem of assessing in situ states; and (iii) explore the formulation of novel data-driven and performance-based procedures in mining geotechnics. Towards these objectives, this project uses an integrated approach that considers data science, laboratory tests, microstructural measurements, field tests, and numerical simulations to bring novel insights to the multiscale response of intermediate materials such as mine tailings considering (i) microstructure signatures; (ii) the role of inherent and induced anisotropy and cyclic loadings; (iii) field-scale system effects; and (iv) data-driven approaches for state inversion. Insights gained from this research will be used to probe the value of data science and formalize its use in unraveling high dimensional interactions of mechanical properties of mine tailings and formulating novel data-driven and performance-based procedures using machine and active learning. This project will allow the PI to establish the foundation for an interdisciplinary field at the convergence of data science, mining geotechnics, and performance-based engineering that can demonstrate to the geotechnical and hazard communities the opportunities of embracing data-driven approaches.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.